Evaluation of a Horticultural Activity Programme for Persons With Psychiatric Illness

This study investigated the impact of a horticultural activity program on several aspects of well-being in individuals with psychiatric illness.

**Intervention:** A horticultural program consisting of 10 sessions delivered over two weeks. The abstract does not specify the exact content or duration of each session, but it involved gardening-related activities.

**Comparator:** Conventional sheltered workshop training. Participants in this control group continued their usual activities within the workshop setting.

**Outcome Measures:**

* **Stress, Anxiety, and Depression:** Measured using the Chinese version of the Depression Anxiety Stress Scale (DASS21).

* **Quality of Life:** Measured using the Personal Wellbeing Index (PWI-C).

* **Work Performance/Behavior:** Measured using the Work Behavior Assessment.

The study included 24 participants with psychiatric illness, recruited using convenience sampling. Two participants from the experimental group dropped out after assignment, resulting in:

**Experimental Group:** 10 participants

**Control Group:** 12 participants

The abstract does not provide further demographic details such as age, specific diagnoses, or duration of illness. The study was conducted in a sheltered workshop setting, implying participants were likely undergoing some form of vocational rehabilitation or support.

Participants were assessed before and immediately after the two-week program using standardized questionnaires:

**Depression Anxiety Stress Scale (DASS21):** This is a 21-item self-report questionnaire designed to measure the severity of symptoms of depression, anxiety, and stress over the past week. It consists of three 7-item subscales for Depression, Anxiety, and Stress. Higher scores indicate greater severity of symptoms. The Chinese version (DASS21-C) was used.

**Personal Wellbeing Index (PWI-C):** This is a self-report measure of subjective quality of life, typically assessing satisfaction with various life domains (e.g., standard of living, health, achievements, relationships, safety, community, future security). Higher scores generally indicate better quality of life. The Chinese version was used.

**Work Behavior Assessment:** The abstract does not specify the exact instrument used for this assessment. It is likely an observational or rating scale used by staff within the sheltered workshop to evaluate participants' work-related behaviors, such as punctuality, task completion, social interaction, or adherence to rules.

This study was a **single-blinded randomized controlled trial (RCT)**.

**Study Design:**

* **Randomized Controlled Trial (RCT):** Participants were randomly assigned to either the experimental group (horticultural program) or the control group (conventional sheltered workshop training). This is a strong design because random assignment helps ensure that, on average, the two groups are similar in all characteristics (known and unknown) at the start of the study. This minimizes confounding variables and increases confidence that any observed differences between groups are due to the intervention, not pre-existing differences.

* **Single-Blinded:** Only the participants were blinded, meaning they were not told whether they were receiving the "active" intervention or the control. The researchers or assessors were likely aware of group assignments. Blinding participants helps reduce the placebo effect (where participants improve simply because they expect to) and demand characteristics (where participants behave in a way they think the researchers expect). However, because the assessors were not blinded, there's a risk of observer bias, where their knowledge of group assignments might subtly influence how they interact with participants or interpret results, especially for subjective measures.

**Randomisation:** Participants were "randomly assigned" to either the experimental or control group. The specific method of randomisation (e.g., coin toss, computer-generated sequence) is not detailed in the abstract. Effective randomisation is crucial for the internal validity of an RCT.

**Washout Periods:** Not applicable, as this was a parallel-group design where participants received only one type of intervention.

**Duration:** The intervention group attended 10 horticultural sessions within a period of 2 weeks. This is a relatively short intervention duration. Assessments were conducted immediately before and after this two-week period.

**Statistical Approach:** The study compared "change scores" (post-intervention score minus pre-intervention score) between the experimental and control groups. This approach helps account for individual baseline differences. A p-value of .01 was reported for the DASS21 change scores, indicating a statistically significant difference.

**What this design can and cannot prove:**

* **Can prove:** Given the RCT design, this study can provide evidence for a causal link between the horticultural activity program and changes in self-reported depression, anxiety, and stress *within this specific population and timeframe*. The randomisation helps establish that the intervention, rather than other factors, likely caused the observed differences.

* **Cannot prove:**

* **Generalizability:** Due to the small sample size (N=22 completed) and convenience sampling (participants were not randomly selected from a larger population), the findings may not be generalizable to all individuals with psychiatric illness or to other settings.

* **Long-term effects:** The short, two-week duration means the study cannot determine if the positive effects on stress, anxiety, and depression are sustained over a longer period.

* **Mechanism of action:** The study does not explore *why* horticultural activities might be beneficial (e.g., physical activity, social interaction, connection with nature, sense of accomplishment).

* **Specific components:** It doesn't identify which specific aspects of the horticultural program were most effective.

* **Effect on work behavior and quality of life:** The study found no significant impact on these outcomes, suggesting the program either doesn't affect them or the study was not sensitive enough to detect an effect.

**Major methodological weaknesses:**

* **Small Sample Size:** With only 10 participants in the experimental group and 12 in the control group, the study has limited statistical power to detect smaller but potentially meaningful effects, and the results are more susceptible to chance findings.

* **Convenience Sampling:** This introduces potential selection bias, as the participants might not be representative of the broader population of individuals with psychiatric illness.

* **Lack of Assessor Blinding:** The assessors knowing group assignments could introduce bias, particularly for subjective self-report measures like DASS21 and PWI-C.

* **Short Duration:** Two weeks is a very brief period to observe robust and lasting changes in complex psychological outcomes.

* **Missing Effect Sizes:** The abstract only reports a p-value for the DASS21, but does not provide the actual magnitude of the change in scores, making it difficult to understand the practical significance of the findings.

**Depression, Anxiety, and Stress (DASS21):** There was a statistically significant difference in the change scores of the DASS21 between the experimental (horticultural program) and control (conventional training) groups (p=.01). This indicates that participants in the horticultural program experienced a greater reduction in self-reported depression, anxiety, and stress symptoms compared to the control group.

* *Note: The abstract does not provide the specific mean change scores or effect sizes (e.g., Cohen's d) for the DASS21, only the p-value.*

**Quality of Life (PWI-C):** There were no statistically significant differences in the change scores of the PWI-C between the two groups. This suggests the horticultural program, in this short duration, did not significantly impact participants' self-reported quality of life compared to the control.

**Work Behavior Assessment:** There were no statistically significant differences in the change scores of the Work Behavior Assessment between the two groups. This indicates the program did not significantly affect work-related behaviors within the study's timeframe.

The abstract states a "significant difference" in DASS21 change scores with a p-value of .01, indicating that the observed reduction in depression, anxiety, and stress in the horticultural group was unlikely due to chance. However, the abstract *does not provide the actual numerical change* in DASS21 scores for either group, nor does it provide an effect size (like Cohen's d).

Therefore, while we know there was a statistically significant positive effect, we cannot translate its magnitude into plain English. We don't know if the average reduction in DASS21 scores was small, moderate, or large in a practical sense. For example, we cannot say if it was equivalent to a certain number of points on the scale, or comparable to the effect of a specific medication or therapy. This is a major limitation for understanding the practical relevance of the finding.

The authors acknowledge that this was a "pilot study" and that the "impact of the programme on work behavior and quality of life will need further exploration."

Additional limitations noted by a critical reader include:

**Small Sample Size:** With only 10 participants in the experimental group and 12 in the control, the study has limited statistical power. This means it might miss real effects (Type II error) or that the significant finding for DASS21 could be an overestimate of the true effect.

**Convenience Sampling:** The participants were not randomly selected from a larger population, which limits the generalizability of the findings to other individuals with psychiatric illness or different settings.

**Short Duration:** A two-week intervention is very brief. It's unclear if the observed benefits are sustainable or if longer exposure would lead to greater or different effects.

**Lack of Specific Effect Sizes:** The absence of mean change scores or effect sizes for the DASS21 makes it impossible to gauge the practical significance or magnitude of the observed improvement. A statistically significant p-value doesn't always mean a clinically meaningful effect.

**Single-Blinded Design:** While participants were blinded, the assessors were likely not. This introduces a risk of observer bias, particularly for self-reported outcomes like DASS21 and PWI-C, where assessor expectations could subtly influence data collection or interpretation.

**Self-Report Bias:** DASS21 and PWI-C are self-report measures, which are susceptible to social desirability bias (participants reporting what they think is expected) or recall bias.

**Lack of Detail on Intervention:** The abstract doesn't describe the specific content, intensity, or duration of the horticultural sessions, making it difficult to replicate the intervention precisely.

**Population Specificity:**